Google acknowledges issues with its AI Overviews feature
Associated Press/Photo by Virginia Mayo, file
![Google acknowledges issues with its AI Overviews feature](https://www4.wng.org/_1500x937_crop_center-center_82_line/Google-logo-6-3.jpg)
Head of Google Search Liz Reid in a Thursday blog post acknowledged the inaccurate results in Google’s new AI Overviews feature. Google announced the AI Overviews feature on May 14. The product is intended to answer complicated search queries by drawing from different sources on the web. However, users noticed that AI Overviews gave them inaccurate or misleading answers to search queries.
How did these issues happen? Google claimed that some of these misleading answers stemmed from algorithms that drew from faulty sources, including message boards and satirical websites. The tech giant noted that some of the misleading answers were due to questions probing what it called information gaps, areas in which the AI Overviews had very few high-quality sources to draw from.
How did Google respond? Reid said that Google had refined the AI Overviews to limit the sources it draws from and is actively removing AI overviews that violate its content policies.
![](https://s3.us-east-1.amazonaws.com/world-website-storage/wng-prod/emailicon.png)
An actual newsletter worth subscribing to instead of just a collection of links. —Adam
Sign up to receive The Sift email newsletter each weekday morning for the latest headlines from WORLD’s breaking news team.
Please wait while we load the latest comments...
Comments
Please register, subscribe, or log in to comment on this article.